Accelerating Product Management with AI

---

Large-language models, or LLMs, have taken the world by storm, and it’s not hard to see why. It is becoming increasingly important for product managers to adapt to this new technology and integrate it into their product workflows to stay competitive. In particular, there are a number of simple use cases where adopting a commercial generative AI tool such as ChatGPT, or the more recent Google Bard into your product management workflow will help accelerate your deliverables.

In this article, we will discuss a few product management use cases for LLMs that boost your productivity and improve the quality of your work. With each use case, we will also provide some sample prompts to get started.

Content Generation

Popular LLMs fall into the broader category of generative AI and, as such, are exceptionally good at content generation. The most immediate benefit of these tools for any professional, including product managers, is to speed up the writing process.

With the right prompt, one can easily generate product document templates, email templates, newsletter templates, or any number of useful starting points for the writing process. A simple prompt such as “Give me a document template for a product requirements document” will quickly generate a skeleton document outlining important sections with sample content and a description of what each section should be used for. This is extremely helpful for quickly writing important documents.

You can further use the LLM to generate sample content for these sections. If you include as much detail as possible in the prompt, you can even get output close to something you can include directly in your document. One word of caution, however, is that with current models available, generated text has a fairly robotic tone and repetitive structure, so it is good to rewrite or modify generated text to better match your own writing voice.

Using LLMs in this way will greatly improve your writing throughput, helping you achieve more in less time.

Brainstorming

LLMs like GPT-4 or Bard are trained on a massive data set of internet text from a vast array of sources. This data set is larger than what any human could individually process. In consolidating this broad set of information into a database that’s easily queried with natural language, LLMs provide the perfect companion for product brainstorming. Asking the language model questions about your business problems or product gives you the ability to cross-check your thinking against the model’s consolidated understanding of the topic.

For example, suppose you are building a complex product like an order management system and want to be competitive in the market. You might start by asking a tool like ChatGPT, What are the top 5 must-have features of a commercial order management product?” to which it would respond like

Here are the top five must-have features of a commercial order management product, summarized:

Order Tracking and Status Updates: Real-time tracking and updates on the status of orders, ensuring transparency and timely communication.

Inventory Management: Integration with inventory databases to provide accurate product availability information, real-time updates, low stock alerts, and seamless order allocation and fulfillment.

Integration with Multiple Sales Channels: Ability to integrate with various sales channels (online marketplaces, physical stores, social media platforms) to consolidate orders and ensure consistent customer experiences.

Automation and Workflow Management: Automation capabilities to streamline tasks such as order entry, payment processing, shipping, and customer communication, reducing errors and increasing efficiency.

Analytics and Reporting: Robust analytics and reporting features to track key performance indicators (KPIs), enabling businesses to identify bottlenecks, measure performance, and make data-driven decisions.

As you see, even with a short and simple query, we get a lot of useful insight about must-have features for the product we are trying to build, and it serves as a perfect jumping off point for designing new features. Furthermore, you can easily continue your brainstorming session by returning to ChatGPT and asking further questions, such as “What are some more advanced features of an order management system considered nice to haves?” or “Which of the features top 5 features listed earlier would have the biggest impact on order volume?” to either build a larger roadmap or fine-tune the prioritization of existing features.

This makes LLMs an extremely useful tool for accelerating idea generation and building impactful product roadmaps.

Document Summary

Being a product manager requires a lot of cross-functional communication. This results in the production of a high number of artifacts documenting everything from requirements to ideas to roadmaps and plans. In a large organization, the sheer number of documents that one needs to read to stay on top of new developments is staggering. This is where LLMs can help us.

ChatGPT, for example, is exceptionally good at parsing and understanding large volumes of text, such as a product requirements document, and with the right prompt, can quickly produce a summary of the text. This is easy to achieve by copying and pasting the document text in quotations into ChatGPT and following up with the prompt “Produce a short summary of the quoted text above.”

You can also use prompts to tune the level of detail that the model will use when summarizing the text. Asking the model, “Can you add a little more detail?” or “Can you be less detailed?” will help you get to the right summary, so you can quickly understand the larger volume of work that is going on. This can also be used to proactively provide summaries of your own artifacts to help accelerate the communication of your own ideas.

There are also a number of tools available leveraging LLMs to summarize or analyze external data sources, such as PDF files or other user generated business content. In particular, at PlanEngine, you can leverage the power of LLMs to automate the process of estimating effort and prioritizing features in your product roadmap by connecting to your product execution data sources, such as Atlassian Jira. Sign up for early access to PlanEngine to unlock efficiencies in your product planning process.

Conclusion

In this article, we discussed a few simple use cases for LLMs in day-to-day product management work. In particular, we looked at how LLMs can help accelerate product brainstorming while also speeding up the process of reading and writing various product documents. Thank you for reading, and I hope you learned something useful for your team.

---